230 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
It is obvious that the quantity of water passing 
through a ring that is hauled from a given depth ata 
given speed, is greater than the volume that would pass 
through the same ring under similar conditions 1f a net 
were attached to it. further, the volume will be depen- 
dent on the following factors: (1) Shape of net; (2) Area 
of filtering tissue; (5) Nature of filtering tissue; (4) Area 
of opening; (5) Speed of pull. 
The first problem to be tackled is the determination 
of the quantity of water that would pass through a unit 
area of any given quality of silk (Miillergaze—-the best 
net material) in unit time, under different pressures. 
It must be observed that this determination of the 
filtration capacity of Miillergaze must be made wnder 
water; the filtration must not take place in air or other- 
wise the results cannot be applied to the nets which are, 
of course, submerged when in action. 
In Hensen’s experiments (see Taf. V, oc. ct.) fresh 
water was led from a reservoir to a cleaning apparatus, 
consisting of a filter box made up of two halves of brass 
separated by a piece of No. 20 Miillergaze. Small 
openings in each half rendered it possible to remove any 
air which collected at the surfaces of the silk. From this 
filter, a tube conducted the water to a strong glass 
spherical vessel, the opening of the tube into the vessel 
being guarded also by Miillergaze. This vessel held a 
certain amount of air in order to reduce jerks in the water 
pressure. 
From it, another tube conducted the water through 
a regulating cock which allowed the experimenter to 
control and regulate the pressure. The water passed 
from the regulator to the actual filtration apparatus, a 
brass cylinder which is permanently closed at one end, 
and which has a circular opening at the other, so con- 
